As memristor-simulating synaptic devices have become available in recent years, the optimization on non-linearity degree (NL, related to adjacent conductance values) is unignorable in the promotion of the learning accuracy of systems. Importantly, based on the theoretical support of the Mott theory and the three partial differential equations, and the model of conductive filaments (CFs), we analyzed and summarized the optimization schemes on the physical structure and the extra stimulus signal from the internal factor and external influence, two aspects, respectively. It is worth noting that we divided the extra stimulus signals into two categories, the combined pulse signal and the feedback pulse signal. The former has an internal logical optimized phenomenon, and the composition of only two parts in each cycle leads to a simple peripheral circuit. The latter can obtain an almost linear NL curve in software stimulation because of its feature in real-time adjustment of signals, but it is complex in hardware implementation. In consideration of space and energy consumption, achieving memristor with different resistive switching (RS) layers can be another optimization scheme. Special attention should be paid to the weaker NL, which could improve learning accuracy at the system level only when the value of other non-ideal properties such as the on/off ratio is within a certain range.

No abstract

Resistive switching characteristics in Cu/ZnO/AZO (Al-doped ZnO) were investigated. Reproducible bipolar resistance switching properties were observed in the single oxide layer (SL)-based device. To improve the switching performance, a CuO–ZnO bilayer (BL) was used to form a Cu/CuO/ZnO/AZO structure. RS characteristics such as retention time, endurance, variations of threshold voltage as well as distribution of resistance were investigated. The results demonstrated that the BL devices exhibit more excellent switching performance than SL devices. The conduction mechanisms of high and low resistance states can be explained by trap-controlled space charge limited current (SCLC) and Ohmic's Law, respectively. The CuO layer is proposed as a “reservoir” of oxygen ions in set process and acting as an oxygen ions “supplier” in reset process, which plays a critical role in recovery/rupture of filament paths and greatly improves the switching characteristics of the device.

This paper proposes a triangle carrier based discontinuous PWM (TCB-DPWM) implement method for threelevel neutral-point-clamped (3L-NPC) inverters. The function equivalent relationship of proposed TCB-DPWM and space vector based discontinuous PWM (SVB-DPWM) is mathematically analyzed in this paper. The proposed TCB-DPWM method effectively simplified the implementation of the SVB-DPWM by injecting twice common mode voltages (CMVs) to original modulation signals, where different DPWM methods can be easily obtained in a unified scheme by setting different values of proportional allocation factor for small vectors in different sectors. In addition, a TCB-DPWM based neutral point voltage balancing method is investigated to demonstrate its application. The polarity of the CMV injection is regulated according to the polarity of the proportional allocation factor to realize the neutral point voltage balancing. Finally, the proposed TCB-DPWM and neutral point voltage balancing method are both verified by simulation and experimental results. The results show that the proposed TCB-DPWM can effectively simplify the implementation of modulation and improve system efficiency. Index Terms-Three-level neutral-point-clamped inverter, voltage source inverter, discontinuous PWM, neutral point voltage balancing.